This study evaluated the effectiveness of proxy voice intervention, embedded within couple enactments, on client-perceived softening. During enactments, direct couple interaction is the focus while the therapist coaches from the periphery. In the context of an enactment, the therapist may use proxy voice when partners appear to be distressed and expressing themselves in terms of secondary emotions by modeling appropriate attachment and self-concept expression. The primary research question was whether therapist use of proxy voice in an enactment would be more likely to bring about softening effects, or whether use of proxy voice was counter-intuitive to enactment conceptualization and would bring about effects related to struggle (e.g. withdrawal or negativity). The review of literature sets forth (1) enactments as common factors; (2) enactments conceptually and operationally; (3) proxy voice in the context of enactments; and (4) the effects of proxy voice on softening versus withdrawal or negativity. Proxy voice occurred 42 times in nine research sessions where proxy voice was delivered repeatedly in a 20-30 minute enactment episode. Results indicated that proxy voice was significantly (both statistically and clinically) associated to softening while dissimilarly linked with withdrawal or negativity. Results also suggested that proxy voice may be used to dampen volatility and foster couple softening during enactment in the following ways: (1) proxy voice temporarily increases the structure of the couple interaction, thereby allowing the therapist to dampen reactivity and model healthy expression before returning to direct couple interaction; (2) proxy voice is a hypothesis of softer emotions that fits the clients' experiences, helps them to feel validated, and encourages them to consider something in a newer, softer way; and (3) proxy voice taps into foundational relationship dynamics surrounding self-concept and attachment experiences that "propel" interaction processes but remain outside conscious awareness or explicit expression for the couple. These preliminary findings suggest that proxy voice intervention embedded within a fluid, carefully delineated, and discriminating model of enactments effectively facilitates essential elements of couple interaction (expression of primary affect, and self-concept and attachment threats) while promoting self-reliant couple interaction and increased couple softening.